[unreadable] Functional magnetic resonance imaging (fMRI) of the brain, based upon Blood Oxygen Level Dependent (BOLD) contrast, produces four-dimensional data sets (three spatial dimensions and one time dimension), containing information relating to the brain activity during task performance. Development of signal processing methodologies for fMRI data remains in its infancy. Presently available methods used for construction of brain activation maps from fMRI data do not make full use of the entire space-time information content of the image data. Many applications await better understanding and exploitation of the extensive spatiotemporal information intrinsic to fMRI data sets. The proposed study will investigate the use of the Space-Time Adaptive Processing (STAP) algorithm in post-processing of fMRI data to elucidate spatial and temporal connectivity of cortical activation pathways. The STAP algorithm we propose for fMRI analysis is adapted from the model originally developed for radar signal processing and, unlike current fMRI processing methods, processes the entire space-time data as a single spatiotemporal set, thus preserving spatial relationships. Initial studies have demonstrated the potential of STAP when applied on a small scale to multidimensional fMRI data. The results indicate that the STAP filter exhibits a high degree of accuracy in detecting small changes in signal intensity in both space and frequency. In the proposed study, a more complete model will be developed to establish its efficacy when applied to larger and more varied fMRI data sets. The new model will improve on the existing small scale model not only in spatial and temporal extent but also in utilization of the phase information, which may provide additional insight into temporal connectivity. Algorithms for both fully adaptive and partially adaptive STAP will be attempted. Fully adaptive STAP requires the solution of a system of linear equations of size MN, where M is the number of sensor elements and N is the number of time samples. Partially adaptive schemes such as element space and beamspace reduce the dimensionality of the problem. [unreadable] [unreadable]